A brief history of the global tower market

global-bible-feature.png

TowerXchange summarises the current state, and origins of, the independent tower market worldwide

This article introduces the various tower company business models, sizes the market, and contrasts the differences in the tower industry from continent to continent. We conclude with a look at the future evolution of tower companies, distinguishing between vertical real estate ‘purists’ and diversified communications infrastructure providers.

Once upon a time, telecom towers were depreciating assets, built to serve the needs of one mobile network operator (MNO), occasionally (often reluctantly) shared on a site for site ‘swap’ basis. Even today, the management of 1,460,581 operator-captive towers is not a core, value-creating business for those MNOs.

Two thirds of the world’s 4,412,963 towers (a total of 2,952,382) have now been transferred from MNOs to tower companies (towercos), or built by those towercos. Towercos are specialist telecommunications real estate sales and engineering firms that have professionalised the management of towers and rooftop sites, and added value by leasing capacity on those towers to multiple tenants, the majority of which are MNOs. Towercos have deployed improvement capex to improve efficiency; they have brought to bear operational excellence and standardisation programmes; they have accelerated rollouts; and in simple terms they’ve put more tenants on the towers, creating value and efficiencies.

One of the key metrics for a towerco is the tenancy ratio, most simply defined as the number of tenants on a portfolio of towers divided by the number of towers in that portfolio. TowerXchange believe the average tenancy ratio of towerco owned towers to be 1.62x globally, compared to less than 1.1x on MNO-captive towers. If you exclude the effect of China Tower’s huge but relatively recently marketed portfolio, which has a 1.49x tenancy ratio but which is being leased up as fast as any towers in the world, the average tenancy ratio on towerco towers rises to 1.95x.

The creation of tower companies has inaugurated an era of tower sharing. Since American Tower, SBA Communications and Crown Castle led the revolution in the mid-1990s, in a little over 20 years, towercos have created a US$276bn, highly investible global infrastructure asset class, with valuations outperforming just about every comp.

There are currently 20 listed towercos worldwide, but the industry’s valuation index remains dominated by the spectacular performance of the three giant U.S. publics (American Tower, Crown Castle and SBA Communications). The most recent addition to the roster of listed tower companies, China Tower Corporation, was the world’s largest IPO since 2016 when it listed in August 2018, at the time valuing the company at US$27.6bn.

The 259 privately owned towercos and joint venture infracos tracked by TowerXchange own a total of 579,349 towers between them.

The world’s 20 listed towercos

20-listed-towercos.png

Figure one: TowerXchange is tracking 276 towercos that now own 66.9% of the world’s 4.4mn towers

figure-one-towercos.png

As shown in figure one, the tower industry is a top-heavy ecosystem: 56.5% of the world’s towers are on the balance sheets of the dozen largest towercos, each of which has a portfolio of over 20,000 towers. A fragmented ecosystem of 29 further towercos own between 5,000-20,000 towers, and a “long tail” of several hundred sub-5,000 tower towercos, together own 10.4% of the world’s towers.

The tower industry is both fragmented and diverse, with a range of business models, lease rates and contractual terms, particularly in that “long tail” of smaller, mostly privately owned towercos, which creates a challenge for larger towercos seeking to consolidate the smaller players.

With the sole exception of North and Eastern Asia (excluding China), towercos are now found worldwide. But not all towercos are based on the same business model blueprint first conceived in the U.S. in the mid 1990s. The contractual terms, particularly the fixed escalators in a low interest economy, negotiated in the U.S. by American Tower, Crown Castle, SBA Communications and their peers created a “gold standard” of investible contracts that may never again be replicated. Nonetheless the towerco blueprint has been versioned to meet the needs of different operators, different market dynamics, and different regulatory regimes.

TowerXchange recognises three simplified sub-categories of towercos, as shown in figure two. The “pureplay independent towercos” category most closely follows the original U.S. blueprint from the mid 1990s – they are public or privately owned towercos with little or no residual equity retained by MNOs. 242 pureplay independent towercos identified by TowerXchange, exemplified by American Tower, Cellnex and Protelindo, own 563,024 (12.8%) of the world’s 4.4mn towers, as shown in red in figure two.

Figure two: Tower ownership by business model, regional comparison

figure-two-business-model.png

One development that could move the needle in terms of tower ownership worldwide is incoming CEO Nick Read hinting that Vodafone’s hitherto-retained portfolio of 55,000 towers could be sold in an effort to reduce the MNO’s $31bn debt (debt reduction being the most common driver for MNOs to sell and leaseback their towers). Whether Vodafone chooses to sell and leaseback their towers, or carve-out and retain control of an “operator-led” towerco, remains to be seem.

The orange segment in figure two represents “operator-led towercos”; towercos that are themselves at least 51% owned by their parent MNO or MNOs. 26 operator-led towercos own 2,323,968 (52.7%) of the world’s towers. That statistic is distorted by the sheer scale of China Tower Corporation – if you extract China Tower and their 1.92mn towers, the percentage of operator-led towerco owned towers drops to 9.1%. However, the fact remains that the carve-out and retention of operator-led towercos is a trend that is gaining momentum. The most recently created operator-led carve out towercos include Altice’s carve-out of SFR TowerCo in France and Towers of Portugal, and Telkom carving out Gyro Towers in South Africa, while TowerXchange understand Viettel are considering carving out their 40,000 towers in Vietnam, potentially as a precursor to extending a carve-out strategy worldwide.

Operator-led towercos often utilise business models and contract structures calibrated to more equitably share value between towerco landlord and operator tenant. For example some operator-led towercos discount the anchor tenant’s lease rate when a second tenant co-locates, others charge a nominal “loading” fee for the overlay of next generation network equipment on a tower, as opposed to charging an additional tenancy, generating what pureplay independent towercos would call “amendment revenue”.

The differences in business model between pureplay independent towercos and operator-led towercos are reflected in valuations, with healthy pureplay independent towercos often valued at EBITDA multiples between the late teens and mid twenties, whereas operator-led towercos have to date been valued in the 7-14x range.

While tenants of pureplay independent towercos often seek the more favorable terms offered by operator-led towercos, their protests will typically fall on deaf ears: if MNOs have been paid a premium for their towers by a pureplay independent towerco, they are going to need to pay a premium to lease them back. It should be noted that MNOs can choose to reduce their total cost of network ownership when selling towers to a pureplay independent towerco – they just won’t release as much cash up front.

Going back to figure two, the third category is joint venture infracos, shown in green, which is where two or more MNOs pool their passive (and sometimes active) equipment into a third party company which either manages or owns the assets. 12 joint venture infracos own or manage 65,400 towers, 1.5% of the global total.

Joint venture infracos aren’t strictly towercos, if for no other reason than they seldom proactively market their sites for co-location beyond the joint venture partners, but they are very much part of the TowerXchange community and require their own category in the ecosystem. Until recently joint venture infracos were a phenomenon unique to Europe, specifically the UK, Scandinavia, Poland, Romania and Greece, but MCI, RighTel and Fanasia are in the process of creating the first joint venture infraco in MENA: Iranian Towers.

The origination, replication and consolidation of the tower industry in the Americas

The ‘original’ pureplay independent towerco business model emerged in the mid 1990s in the USA and Canada, where towercos now own 134,004, or over 73%, of the two country’s towers (representing over 80% penetration in the U.S., with low penetration in Canada). Led by American Tower, Crown Castle and SBA Communications, towercos have built and bought the majority of towers and rooftops sites in the U.S., leaving the country’s MNOs uniquely dependent on their networks, which in turn puts the towercos in a relatively strong bargaining position when it comes to lease renewals.

TowerXchange doesn’t cover the North American market, but we do cover Central and Latin America (CALA), which was the first region to which the U.S. towerco model was exported.

Towercos own 52.7% of the towers in CALA, led by American Tower and SBA Communications, more recently joined by America Movil and Telefónica’s operator-led towercos Telesites and Telxius respectively. TowerXchange are tracking a further 28 of the dozens of private towercos in CALA, who between them own 14% of the region’s tower assets, but who build almost half of CALA’s new towers.

A distinction should be drawn between the Latin American and the 12,857 tower Central American market, the latter almost functioning as an extension of the U.S. ‘gold standard’ market as a result of the use of standard contracts, U.S. dollar currencies, and lower-than-perceived country risk, combining to insulate Central American tower companies against the principle risks of international expansion. In contrast, most tower markets in Latin America have been subject to greater political, economic and regulatory uncertainty, and many have been affected by currency devaluations against the U.S. dollar.

While long-term investors remain satisfied with their investments in Latin America, shorter term investors have had a tougher time of late. CALA’s forecast new tower build volumes seem perpetually deferred by MNO consolidation and bankruptcies, and by changing governments and changing policy. The simple fact remains: CALA needs a lot more towers, and the current build volume of ~5,000 new towers per year (the vast majority of which are built by towercos rather than self-deployed by MNOs) will need to more than treble to densify the networks for 5G.

The dynamics of the CALA (and indeed the North American) tower market exemplify consolidation between towercos. Several serial tower entrepreneurs have created ‘build-to-flip’ towercos that specialise in building towers in particular localities or for particular clients, with intent to sell those assets on to a consolidator. This in turn feeds consolidator, or ‘rollup’, towercos that specialise both in acquiring and integrating privately built towers, and in ‘cleaning up’ portfolios of sub-optimum towers, for example those with incomplete sets of permits. The phenomenon of build-to-flip and rollup towercos is most common in the Americas, but is not unique to the region: we’ve seen towers built and flipped in Indonesia, Myanmar and Nigeria, and the trend may soon take effect in China.

For a deeper dive into the tower industry in Central and Latin America, check out TowerXchange’s summary overview of the CALA tower market.

Pureplay independent towercos own just 12.1% of Europe’s slides

The new tower market emerging in MENA notwithstanding, Europe is the tower industry’s least mature market. Defining Europe as inclusive of Russia, the CIS and Turkey, operator-led towercos own 17.6% of the region’s towers, exemplified by Deutsche Telekom’s Deutsche Funkturm, TIM’s INWIT, Telefónica’s Telxius and Altice’s newly carved-out SFR TowerCo and Towers of Portugal. There is also a unique prevalence of joint venture infracos, 12 joint venture infracos operate (but don’t necessarily own) around 64,400 of Europe’s towers (10.7%). The fact that pureplay independent towercos have been able to get their hands on just 12.1% of Europe’s towers is illustrative of two things. One, many European carriers remain anxious to retain control of their networks. And two, there is a perception that the maturity of European tower networks means that many markets will generate only ‘infrastructure-type’ returns – barely into double digits. This in turn explains the increased involvement of infrastructure funds, such as KKR and PGGM, in European tower investment.

While there are 20 towercos with more than 1,000 towers in Europe, and 27 smaller towercos, the phenomenon of the build-to-flip towerco is rare as build volumes are currently low and most MNOs keep their search rings to themselves.

As a relatively mature 3G-4G market, with considerable overlap in networks, for the last five years Europe has demonstrated how organic growth is almost entirely cancelled out by decommissioning in an ‘overbuilt’ market. However, 5G is changing that. A recent TowerXchange survey of the 16 largest and most investible tower markets in Europe forecast that 51,500 new towers would be built by 2023. Factoring in the rest of Europe, the total tower stock in Europe may increase by 10% in the next five years, driven by densification for 5G, which will in turn prompt the deployment of tens of thousands of small cells.

For more detail on the European tower industry, please read TowerXchange’s analysis of the European tower industry.

Asia’s unique tower industry

India was the birthplace of Asia’s tower industry in 2008. While the Indian tower industry was inaugurated by a pureplay independent towerco that became Viom Networks (ultimately acquired by American Tower), India’s relatively uniform and MNO-friendly contractual norms and lease rates have been largely defined by operator-led towerco giants Indus Towers and Bharti Infratel, who have themselves announced intent to merge in 2019. With relatively low lease rates, and other favorable terms including discounts for anchor tenants when additional tenants are added to a tower, the efficiencies unlocked by India’s towercos accelerated the country’s mobile rollout, and contributed to optimising costs.

The key theme in India’s tower industry today is the impact of MNO consolidation. Whereas Indian towers were once changing hands for over US$120,000 each, since 2015 valuations have been in the US$55-70,000 per tower range. The reason is simple: where once India teemed with over a hundred regional MNOs, and more recently nine MNOs with nationwide aspirations, now MNO consolidation seems likely to result in 4-5 national operators, headlined by the merger of #2 and #3 operators Vodafone and Idea Cellular. Tens of thousands of tenancies will not be renewed, while battles continue to recover contracted lease revenue from ailing operators.

The stimulus for India’s market restructuring was innovative new entrant MNO Reliance Jio jacking the Indian mobile and tower industries into the 4G era, grabbing huge market share, and making voice free. With Jio heavily dependent on data, fiberisation was critical, yet many of India’s independent towers were not connected to fibre, hence Jio has more than half its network on small cells and single tenant proprietary lampposts. Jio has also acquired the infrastructure assets from its elder sibling RCOM, whose wireless business is winding up. Jio has acquired spectrum as well as substantial fibre holdings, along with towerco Reliance Infratel (~43,000 towers, on which Jio is a tenant on approximately 30,000 sites, and which TowerXchange cautiously still counts as a towerco, pending reconfirming whether the assets will still be leased to third parties on a non-discriminate basis).

Let’s look at China. In 2014, with the approval of the General Office of the State Council and led by State-owned Assets Supervision and Administration Commission of the State Council (SASAC) and the Ministry of Industry and Information Technology (MIIT), China’s operator-led towerco giant China Tower Corporation (CTC) was formed to promote a culture of infrastructure sharing, also referred to as “co-build, co-share.” After an initial ~1.44mn existing China Mobile, China Telecom and China Unicom towers were injected in 2015, CTC has built a mind-boggling 480,000 towers in three years, driving tenancy ratios from an initial 1.1x to 1.49x, and saving tens of billions of RMB in the process. CTC’s journey culminated in the world’s largest IPO since 2016 when the company floated on the Hong Kong Stock Exchange in August 2018, selling a 25% stake for US$6.9bn.

Alongside CTC, China is teeming with independent towercos – there are at least 200 privately owned tower companies in China, owing around 50,000 towers between them. The top ten independent towercos in China own ~80% of the independent towers, led by Guodong and Miteno with ~20,000 and ~6,000 towers respectively.

The mature Indonesian tower market stands apart from the rest of Asia as it features several pureplay independent towercos of considerable scale, and they have contributed to a relatively commercial, Americas-like towerco business model. The Indonesian market is led by Protelindo, Tower Bersama and STP, who have achieved substantial growth through buy and leaseback deals with the country’s MNOs, by rolling up smaller private towercos, and through steady organic growth, which offsets lease up, keeping tenancy ratios at around 1.7x. Telkom’s Mitratel is emerging as a strong operator-led towerco, while a long tail of 30+ privately owned local independent towercos remain acquisition targets.

Indonesia is a slightly more mature mobile market than many others in Southeast Asia; ARPUs are higher and 4G rollouts are further progressed, which means IBS, fibre and small cells are land grabs as much as are macro towers. This diversification of the towerco inventory beyond ground based towers and rooftops is exemplified by the acquisition of technical innovators iForte by Protelindo, and Bit by STP, as well as partnerships such as those of smaller towercos Balitower with Jakarta to operate their CCTV network in return for offering the poles as small cell sites, or towerco Pekape’s partnership with retail giant Alfamart.

Beyond China, India and Indonesia, the rest of Southern and Southeast Asia is primarily populated by single country towercos, with the exception of edotco, in six countries, and OCK in three. edotco merit special mention as one of the world’s most accomplished operator-led towercos. Initially carved out of Axiata, edotco has seen its parent company’s stake diluted to 62.4% by investments from INCJ, Khazanah and KWAP, demonstrating edotco’s commitment to genuine independence. edotco is believed to be scaling toward an IPO in perhaps 12-18 months.

For a more detailed guide to the tower market in Asia, visit TowerXchange’s summary overview of the Asian tower market.

Towercos evolve into powercos in Africa

The ‘Big Four’ towercos (IHS Towers with 22,860 towers in SSA, American Tower with 11,098, Helios Towers with 6,533 and Eaton Towers with over 5,000) own 26.8% of the towers in Sub-Saharan Africa, representing the majority of the sites in investible markets with investible anchor tenants. This explains why the pipeline of sale and leasebacks in Africa has slowed and the growth of the region’s towercos plateaued. Operator-led towercos, and a growing band of emerging smaller towercos, bring the total penetration of towercos in SSA to 38.6%. Special mention should be made of Atlas Towers, which passed the 700 tower milestone in September 2018, and which remains on track to have over 1,000 towers by the end of 2019. Atlas Towers are adding new towers to their portfolio as fast as anyone in Africa.

The African towerco business model evolved to solve the continents’ MNOs’ number one challenge: energy. By transferring both tower and power assets to their towerco partners, Africa’s MNOs have been able to refocus on their core business, while the towercos have cultivated centres of excellence in power management, improving uptime and efficiency. This focus on energy efficiency has also prompted the emergence of a parallel business model to the towerco, that of an Energy Services Company, or ESCO. ESCOs already own or operate the power systems at 15,880 SSA cell sites, and TowerXchange forecasts this total will rise to 43,800 by 2024. TowerXchange recently published a detailed study of the global telecom ESCO market, which you can read here.

For a deeper dive into the tower market of SSA, read our market overview here.

Tower industry finally taking root in MENA

While towercos currently own 0.5% of the 230,189 towers in the Middle East and North Africa (MENA), the region’s first landmark sale and leaseback is imminent with IHS Towers expected to close the acquisition of 1,600 Zain Kuwait sites around the time this article goes to press. IHS has also entered into exclusive negotiations with Zain to acquire a further 8,000 sites in Saudi Arabia. Another tower sale and leaseback process will commence imminently in MENA, bringing 4-5,000 towers and rooftops to market.

The joint venture infraco model is also taking hold in MENA, with Iran’s leading and #3 MNOs MCI and Rightel partnering with domestic towerco Fanasia to form Iranian Towers. However, STCs dialogues to form a joint venture with its peers in Saudi Arabia have faltered.

Egypt is also an important MENA market, with substantial new build contracts potentially up for grabs by towercos, and Orange in the latter stages of an ESCO RFP.

For more information on the emerging MENA tower market, see our market overview here.

Figure three: Africa’s towerco-owned towers

 

figure-three-africa.png
The future of the tower industry: a bifurcation between vertical real estate specialists and providers of diverse yet complementary communications infrastructure

Figure four represents the vertical real estate-centric world as most towercos, and their investors, know it today. To use a baseball analogy, this word cloud is full of good pitches to hit: towercos are in their comfort zone, and the opportunities and issues illustrated on this slide are in their wheelhouse. The tower industry as we know it today is ultimately a simple business: we build towers, we buy towers, we lease them up: we add value.

Figure four: Tower industry 1.0: Vertical real estate

figure-four-tower-industry.png

But with towercos now owning two out of three of the world’s towers, and many remaining towers stranded in uninvestible markets, towercos are running out of inorganic growth opportunities. And with an increasing proportion of new sites being infill city poles and small cells, the ground based tower rollout is maturing and growth is plateauing. How do we extend our industry’s growth narrative? Firstly, we must continue to open up virgin territories; from Argentina to Zambia. When towers are ‘trapped’ in uninvestible markets, for example where Foreign Direct Investment is capped at 49% or less, can our investment criteria be relaxed? Alternatively, do we need a global body to lobby regulatory stakeholders to make more new tower markets investible? The International Digital Infrastructure Alliance has been formed specifically for this purpose – read about the IDIA here.

Secondly, can we expand our horizons to provision a broader set of connectivity solutions, and to engage with a broader range of alternate site and asset typologies?

Figure five shows ‘The Future Network’ illustrated through a new word cloud which includes many concepts representing pitches outside the tower industry’s metaphorical strike zone. But you can still hit a home run on a pitch outside the zone!

Figure five: The future network

figure-5-the-future-network.png

We’re moving into an era when a cell site is increasingly not a macro tower. And we’re moving to an era when network topographies will be redefined by software, not by the exchange of equipment on a tower. A more dynamic market in which carriers demand that their towercos be genuine partners, not landlords. From the threat of disruptive new market entrants leveraging MESH networks of small cells to create high quality coverage for high value customers – making minimal use of macro infrastructure; to the opportunity of converting towers at the network edge into micro data centres, towercos may need to consider doing more than passively selling vertical real estate.

If we’re to transform the tower industry from passive real estate sales agents into strategic partners of carriers, we need both a new mindset and a new toolset. Business intelligence tools like Crowd SiteIntel from M2 Catalyst, which uses analyses of crowd sourced data to transform every subscriber into a human driver tester, can enable towercos to view your sites through the lens of the network planner, enabling the proactive promotion of co-location and build to suit services as solutions to carriers’ problems. Such tools can also help focus towercos’ capital deployment into fibre, IBS and small cells, while also providing a reality check for lease-up forecasts in due diligence.

Conclusions – the shape of towercos and wireless networks to come

The future of the tower industry worldwide will be defined by innovation, consolidation and bifurcation.

In terms of bifurcation, towercos will increasingly be grouped into two categories: those who strictly adhere to the blueprint of a REIT towerco, who exhibit steady growth, who (eventually) pay good dividends, and who build and buy good ground based towers, with good paper, in good locations. And there’s nothing wrong with that model – it generates proven returns, and risks are controlled.

Then there will be towercos that will diversify beyond the tower industry 1.0 blueprint; leveraging street furniture for urban infill, pushing beyond DAS in landmark buildings to deploy small cells in ‘middle market’ premises, laying and acquiring fiber, and eventually converting selected cell sites into micro data centres. Early adopters like Digital Bridge and Crown Castle have already planted their flags in this innovative second category.

TowerXchange are not advocating one path – adherence to the existing blueprint, or diversification as a ‘Future Network Provider’ – but we do feel towercos should either decide to diversify, or decide to focus primarily on their core business.


Tower count methodology

TowerXchange collects data from the public disclosures of 20 listed towercos, combined with informal disclosures from the management teams of several hundred private towercos. While we are asking for counts of finished ground based towers, and alternate site typologies (including rooftop poles) with capacity for multiple tenants, one person’s definition of a multi-tenant suitable site is not the same as the next. Finally, we also include an estimated 30,500 towers owned by pureplay independent towercos but which TowerXchange have not yet identified. The vast majority of these unidentified towers are in China, Indonesia and Vietnam, and their existence and estimated counts have been confirmed by local trade associations.


 

Gift this article